Buscador UdeSA

Filtrar búsqueda por categorías

A local depth based clustering procedure

Fernández-Piana, L., & Svarc, M. (2026). 
In this paper, we introduce a two-stage partitioning clustering procedure based on local depths. In the first stage, we find clusters of the local depth inner region of level In the second stage, the remaining points are assigned to one of these clusters according to a proximity criterion. In this way, the clusters found in the first stage play the role of flexible centers, that aim to mimic the shape of the groups. We analyze the performance of the procedure on multivariate and multivariate functional data, on real and synthetic datasets showing remarkable results.
A local depth based clustering procedure